Predictors of noncompliance with warfarin therapy in an outpatient anticoagulation clinic

Study Objective. To test variables associated with the Health Belief Model, the Patient Satisfaction Model, and patient demographics in predicting compliance with warfarm therapy. Design. Descriptive study using chart and pharmacy database review as well as patient questionnaires. Setting. Outpatient anticoagulation clinic in an academic medical center. Patients. Seventy-five patients who had received warfarin therapy over the previous 6 months. Measurements and Main Results. Clinicians interviewed patients using a subscale of the Beliefs About Anticoagulation Survey and a Patient Satisfaction Model. Perceived barriers to compliance with warfarin therapy and marital status explained 21.7% of the variance in calculations of refill adherence (p<0.001). Compliance with warfarin therapy was assessed using a validated compliance self-report survey and outpatient pharmacy dispensing data. Perceived barriers and current living conditions explained 33.2% of the variance in self-reported noncompliance (p<0.001). Logistic regression analysis, conducted to attempt to predict patients with an 80% refill adherence threshold, correctly classified compliance 85.3% of the time. Conclusion. This investigation provided some insight into the factors predictive of noncompliance in a low-income, outpatient population. Perceived barriers, marital status, living arrangements, and drug regimen played significant roles in warfarin noncompliance. Our results provide more evidence supporting the use of refill adherence and patient self-report as measures of noncompliance. This linkage helps validate compliance as a useful surrogate of patient health outcomes. This study also offers a model that can help clinicians identify patients at significant risk of noncompliant behavior.